Lung cancer screening with MRI: results of the first screening round

J Cancer Res Clin Oncol. 2018 Jan;144(1):117-125. doi: 10.1007/s00432-017-2521-4. Epub 2017 Sep 20.


Purpose: To evaluate the suitability of MRI for lung cancer screening in a high-risk population.

Materials and methods: A 5-year lung cancer screening program comparing MRI and low-dose CT (LDCT) in a high-risk population was initiated. 224 subjects were examined with MRI and LDCT. Acquired MRI sequences were T2w MultiVane XD, balanced steady-state-free precession, 3D T1w GRE, and DWI with a maximum in-room-time of 20 min. Categorization and management of nodules were based on Lung-RADS. MRI findings were correlated with LDCT as a reference. Here, we report on the first screening round.

Results: MRI accurately detected 61 of 88 nodules 4-5 mm, 20 of 21 nodules 6-7 mm, 12 of 12 nodules 8-14 mm, 4 of 4 nodules ≥ 15 mm (solid nodules), and 8 of 11 subsolid nodules. Sensitivity/specificity of MRI for nodule detection was 69.3/96.4% for 4-5 mm, 95.2/99.6% for 6-7 mm, 100/99.6% for 8-14 mm, 100/100% for ≥ 15 mm (solid nodules), and 72.7/99.2% for subsolid nodules. The early recall rate was 13.8% for MRI and 12.5% for LDCT. Following Lung-RADS recommendations and based on interdisciplinary consensus, histology was obtained in eight subjects. The biopsy rate was 3.6% for MRI and 3.4% for LDCT. In all of these eight cases, the nodules were carcinomas, and all of them were accurately detected by MRI.

Conclusion: The results of the first screening round suggest that MRI is suitable for lung cancer screening with an excellent sensitivity and specificity for nodules ≥ 6 mm.

Keywords: Lung cancer; MRI; Screening.

Publication types

  • Comparative Study

MeSH terms

  • Early Detection of Cancer / methods
  • Humans
  • Lung Neoplasms / diagnostic imaging*
  • Lung Neoplasms / pathology
  • Magnetic Resonance Imaging / instrumentation
  • Magnetic Resonance Imaging / methods
  • Middle Aged
  • Multimodal Imaging / methods
  • Multimodal Imaging / mortality
  • Tomography, X-Ray Computed / instrumentation
  • Tomography, X-Ray Computed / methods